Roads: The Key to Efficient Terraforming

[Earthmichael]

Many people only build roads where they actually need to travel, not realizing the importance of roads to terraforming in general.

Most people build 1 speed formers, due to the high cost of 2 speed formers.  So a former needs to waste a turn entering a square to be terraformed, and then start terraforming next turn.

This is where roads enter the picture.  Even if a former has moved 2 squares along a road, the 1/3 movement point left is enough to get a full turn of terraforming done.

So my strategy is to send some advance terraformer units building roads.  You can even build a couple of the expensive 2 speed formers for this, since they are dedicated road-building terraformers.

Then I create packs of terraformers that follow the roads, doing whatever terraforming I want.  For example, I may have 4 terraformers that are going around planting forests.  As long as they are moving on road, at most 2 squares from their previous location, they can build a forest EVERY TURN!

You want to be careful about plotting the route of your terraforming groups, to make sure that they can be productive every turn.  If you do this, you will have a tremendous terraforming advantage over the players who only build roads as actual highways.  You can have packs of formers productive every turn, while their formers packs have a one turn gap between terraforming each square, as each former spends a turn moving into the next square to terraform.

There is an exploit that can help terraformers (that do not use the road strategy I have outlined) be a little more effective.  But this exploit is banned by most players, and is explicitly banned for games hosted by this site.  See number 10 at
http://alphacentauri2.info/index.php?topic=1519.msg4448#msg4448

The good news is that you can be MORE efficient than this exploit using the road strategy, and that this exploit will not work when using the road strategy.  (You can test this for yourself.  Move 4 terraformers 1 or 2 spaces down a road.  Build a forest.  Now try to reactivate any of the 4 terraformers.  They will not reactivate; the exploit will not work.)  Which only makes sense: if you move 1 or 2 spaces down a road and then terraform every turn, you are getting the maximum possible value out of your terraformers.  No exploit will allow you to do better than that, and you can know that you are getting the most effective use of your terraformers, period, and feel good that you are doing it with a perfectly legal strategy!

[Yitzi]

An important note: This is only worth it when you have at least 3 formers terraforming any given square (it breaks even at 2).  In fact, spreading out your formers makes for an even more efficient terraforming strategy, albeit one that gives returns a tiny bit later in the earliest stages.

[Earthmichael]

If you are going to build a road at all, I think it is best done before terraforming (except remove fungus or terraform level).  For example, if you plant a forest first, I believe building a road takes more time than it would on just open ground.

There are many reasons one might have to revisit a square.  It could be that fungus has overtaken.  Or that your needs have changed, and a forest needs to give way to some other development.  Or you may need to send military units to deal with a worm boil, or an enemy invasion.  Or it may be time to improve the road to a tube.  In an case, having the road makes it faster to deal with any of these things, since the unit can move and begin the improvement (or attack) on the same turn.

I also tend to deal with terraformers in packs that can generally complete the desire terraforming in 1 or 2 turns.  It makes the management of their tasks a lot simpler and faster for me, and is much more effective than automating the formers.  So if I am building forests, I generally use a pack of 4, less with WP, of course.

[Yitzi]

If you are going to build a road at all, I think it is best done before terraforming (except remove fungus or terraform level).  For example, if you plant a forest first, I believe building a road takes more time than it would on just open ground.

Yes in the case of planting forests (not sure about fungus), otherwise I don't think it matters unless you're using at least 2 formers on the square.

There are many reasons one might have to revisit a square.  It could be that fungus has overtaken.  Or that your needs have changed, and a forest needs to give way to some other development.  Or you may need to send military units to deal with a worm boil, or an enemy invasion.  Or it may be time to improve the road to a tube.  In an case, having the road makes it faster to deal with any of these things, since the unit can move and begin the improvement (or attack) on the same turn.

And on how many squares will any of those things happen?  If it's substantially less than the number of squares in your entire territory (I'd guess it probably is), then it probably isn't worth having more than a normal road network.  (Also, remember that worm boils and enemies can use roads too; if you build roads everywhere they can generally get where they want easily, but if you have only a network going through cities you'll have a substantial mobility advantage in your territory.)

I also tend to deal with terraformers in packs that can generally complete the desire terraforming in 1 or 2 turns.  It makes the management of their tasks a lot simpler and faster for me, and is much more effective than automating the formers.  So if I am building forests, I generally use a pack of 4, less with WP, of course.

It does, however, add substantially to the cost of terraforming; in the case of forests, you end up spending 6 former-turns (1 to move the road-builder, a second to build the road, and 4 to build the forest) instead of 5, a 20% increase.

[Earthmichael]

It does, however, add substantially to the cost of terraforming; in the case of forests, you end up spending 6 former-turns (1 to move the road-builder, a second to build the road, and 4 to build the forest) instead of 5, a 20% increase

But, I get the road built, which I want anyway.  If I wait to build the road later, I will need to move a former back onto the square, costing an extra trun, and it will take longer to build the road after the forest has been planted.

I also get the value of getting the forest 3 turns sooner than I would otherwise get it.  That is 1/2/1, or at least 0/2/0 (with a crawler), for three turns, just because I built the road and could use a former pack.  And it is much less realtime consumed by me to play the game this way, which is an important consideration since I typically have at least 4 games going at the same time.

[Buster's Uncle]

I do the same for the same reasons - 'though it hurts a little to do so in the opening turns.  Do you pre-road everything, or just a path towards the closest base(s)?

[Earthmichael]

I pre-road pretty much in the order I want to terraform, although I do build roads between bases even if I don't intend to terraform those square for awhile, just for common defense, and ease of movement.

When I get the tech (or if reverse engineering is allowed), I usually build a couple of speed 2 formers dedicated to road construction.

[Yitzi]

But, I get the road built, which I want anyway.  If I wait to build the road later, I will need to move a former back onto the square, costing an extra trun, and it will take longer to build the road after the forest has been planted.

Of course if you're planning to build the road anyway it's better to do it before, but why would you want roads everywhere rather than just a network between bases?  Roads everywhere isn't a big boost in terms of your movement, and is a huge boost to enemy movement (hence, worse defense).

I also get the value of getting the forest 3 turns sooner than I would otherwise get it.

An average of 1 turn before, actually: Instead of getting 5 forests on turn 4 (and then repeating), you get 1 forest on turn 1, 1 forest on turn 2, 1 on turn 3 1 on turn 4, and 1 on turn 5 (and then repeating).  Of course, that's only the first set of 5; the second set of 5 is on average at the same time as with non-roads, and then it goes further from there.

And it is much less realtime consumed by me to play the game this way, which is an important consideration since I typically have at least 4 games going at the same time.

Ah; I prefer to play only one game at a time, and often make it a long one.

[Earthmichael]

I have to revisit a lot of my squares at one time or another, particularly if there is rising water, which there almost always is, even if I keep my ecodamage to zero.

When I want to revisit a square to build a borehole or a mirror or to drill to aquifer or to raise land, I don't want to send a single unit and wait forever for it to happen.  So I have typically packs of 4 or 6 or 8 formers (sometimes more) for these things.  Furthermore, as I move along my road to get to where I want to terraform, my pack moves at most 2 square down the road, and then builds a forest or farm or sensor or something.  This takes me a little bit longer to get to where I want to terraform, but my formers are fully productive all along the way.

[Yitzi]

I have to revisit a lot of my squares at one time or another, particularly if there is rising water, which there almost always is, even if I keep my ecodamage to zero.

Ah...what, the other factions produce a lot of ecodamage?
Although even so, if you can delay it long enough, "launch solar shade" is a perfect counter to rising water if you can get it passed.

When I want to revisit a square to build a borehole or a mirror or to drill to aquifer or to raise land, I don't want to send a single unit and wait forever for it to happen.  So I have typically packs of 4 or 6 or 8 formers (sometimes more) for these things.

But on how many squares do you do that?

Furthermore, as I move along my road to get to where I want to terraform, my pack moves at most 2 square down the road, and then builds a forest or farm or sensor or something.  This takes me a little bit longer to get to where I want to terraform, but my formers are fully productive all along the way.

That should work fairly well even if you just have roads between cities.

[Earthmichael]

I have found that most multiplayer games end up dealing with a lot of rising water, particularly if AI is involved.  Most AI faction seem to pay no attention to ecodamage.  So revisiting square to raise them happens far more often than I would like.

Also, even with low or even zero ecodamage, I still get fungus spread and an occasional fungal pop.  If I have a road already on that square, it is much easier to get a former pack to the square to get rid of the fungus and replace my original terraforming.  Fortunately, speading fungus and pops do not damage the underlying roads or magtubes.

I also do a lot of "daisy-chaining".  For example, when I create a new supply crawler (pre-magtube), and my nearest mine or energy farm is a couple dozen squares away, rather than sending my crawler to the mine and taking whatever delay that involves, instead I can move the crawler 3 squares along the road to a mine, displacing the crawler there, and send that crawler 3 square further down the road to another mine, etc., until the last displaced crawler is on an unused mine or energy farm.  But this requires roads between the various mines and solar farms and such for this to work.

Finally, when I have to deal with an enemy incursion, whether another player or mind worms, I like to have a very active defense.  My primary defense is to send speed 2 or 3 units down the roads/magtubes to attack the enemy units whereever they might be.  If I have a good network of roads, this kind of active defense is very doable. 

If I can catch units when they first land, I can kill one or more units, doing collateral damage to all of the units in the stack, making the rest of the attackers far less effective. 

If I catch a stack of land-based mindworms, the collateral damage kills all of the mindworms in the stack.  This is particular important in the mid to late game, where a single stack of mindworms may have 10 or more units.  If you wait even one turn, all of those units have an opportunity to spread out and attack, and given the 50% land attack bonus, probably each of them killing one of your units.  Furthermore, after they have spread out, each one must be killed individually, instead of killing all land-based mindworms in a single attack.

[Buster's Uncle]

If you can keep track, you can run out the movement points for a crawer, then go back and click on it to get it working the square.  I have to use the technique a lot when I'm trying to reach something distant...

[Earthmichael]

Yes, keep all of your crawlers working every turn, even while moving them to a distant target.  After you have moved your crawlers, whether daisy-chaining as I mentioned earlier, or just straight movement, click on it to work the square you are currently on (unless a city or another crawler is already working that square).  Just look for any crawlers without the "O" symbol, and get them working on something!

[Yitzi]

I have found that most multiplayer games end up dealing with a lot of rising water, particularly if AI is involved.  Most AI faction seem to pay no attention to ecodamage.  So revisiting square to raise them happens far more often than I would like.

If it's that problematic, it's not too difficult (if everyone agrees) to reduce the rate at which the water rises.

Also, even with low or even zero ecodamage, I still get fungus spread and an occasional fungal pop.  If I have a road already on that square, it is much easier to get a former pack to the square to get rid of the fungus and replace my original terraforming.  Fortunately, speading fungus and pops do not damage the underlying roads or magtubes.

Naturally, but "occasionally" isn't going to justify having roads everywhere, especially when you consider that it's a lot weaker defensively.

But this requires roads between the various mines and solar farms and such for this to work.

Actually, it's sufficient to just have roads along the path, so roads between bases would probably be enough.

If I have a good network of roads, this kind of active defense is very doable. 

Definitely; having a network of roads makes sense, it's just having roads everywhere that I'm questioning, and that's actually counterproductive in the "enemy invasion" situation, since a road network can be channeled through your bases (hence being largely unusable by enemies), whereas roads everywhere cannot.

[ete]

Unless bases are really far apart, assuming adequate sensor/ZOC coverage and an active defense, you should not be too worried about land attacks going for your heartland (unless maybe you made every square a magtube, which I don't think he's suggesting). If they have the force to break the first base they hit, they'll take it. If they go further in between your bases trying for a soft target you'll have time to counterattack earlier. I don't see it as likely to be a significant weakness defensively, and even the occasional returning to squares justifies them if you've got former turns you don't urgently need elsewhere.